In conventional distributed computing systems, received messages are added to a queue of a message consumer, and are automatically executed once they reach the head of the queue. There is an implicit assumption in conventional distributed computing systems that versions of software components included in the message consumer correspond to versions of software components included in a message producer. If the versions of software at the message consumer do not correspond to those of the message producer, the message consumer may be unable to process the message.
In some conventional distributed computing systems, the message consumer may generate an error message if the message cannot be executed. The error message may alert an administrator that the message consumer needs to be updated. The administrator must then manually update the message consumer to enable it to process subsequent messages. Conventional distributed computing systems do not include a mechanism for dynamically updating the message consumer prior to processing the message.